[squid-dev] Sad performance trend

Alex Rousskov rousskov at measurement-factory.com
Mon Sep 12 15:00:42 UTC 2016


On 09/12/2016 07:25 AM, Kinkie wrote:
>     > On 27/08/2016 12:32 p.m., Alex Rousskov wrote:
>     >>         W1  W2  W3  W4  W5  W6
>     >>   v3.1  32% 38% 16% 48% 16+ 9%
>     >>   v3.3  23% 31% 14% 42% 15% 8%
>     >>   v3.5  11% 16% 12% 36%  7% 6%
>     >>   v4.0  11% 15%  9% 30% 14% 5%

> would it be possible
> to share these benchmarks with the larger community, 

Not sure what you mean by sharing benchmarks but I can, of course, share
detailed test results and [trivial] Polygraph workloads that were used
for these tests. However, they have no significant value on their own,
and their unpolished specifics are likely to distract us from the core
problem. What value do you see in publishing those dirty details?


> possibly by leveraging the automations we are already using?

Sorry, I do not know what that means. AFAICT, the Project (i.e., "we")
is not using any automated performance testing at this time (or it is so
broken that it does not even detect the obvious trend above). I do not
have enough cycles to fix and maintain that myself, but I have sent the
Foundation Board a specific proposal on how to address that problem.

Alex.



More information about the squid-dev mailing list